Qubes 3.0 rc1 Error reset PCI network card in notebook
jeremia...@web.de
I can not start the sys-net VM with my PCI network card attached.
Internel error:
Unable to reset PCI device 0000:04:00.2:
internal error: Active 0000:04:00.0 devices on bus with 0000:04:00.2
not doing bus reset
This happens only in Qubes 3.0 rc1 and not in Qubes 2.0
Network card:
Ethernet controller: Realtek RTL 8111/8168/8411 PCI Express Gigabit
Best regards
J. Eppler
Marek Marczykowski-Górecki
Hash: SHA1
On Tue, Jun 02, 2015 at 11:20:22PM +0200, jeremia...@web.de wrote:
> Hello,
>
> I can not start the sys-net VM with my PCI network card attached.
>
> Internel error:
> Unable to reset PCI device 0000:04:00.2:
> internal error: Active 0000:04:00.0 devices on bus with 0000:04:00.2
> not doing bus reset
>
> This happens only in Qubes 3.0 rc1 and not in Qubes 2.0
all of them to sys-net. Check for devices 04:00.0 and probably 04:00.1.
> Network card:
> Ethernet controller: Realtek RTL 8111/8168/8411 PCI Express Gigabit
>
> Best regards
> J. Eppler
>
Best Regards,
Marek Marczykowski-Górecki
Invisible Things Lab
A: Because it messes up the order in which people normally read text.
Q: Why is top-posting such a bad thing?
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1
iQEcBAEBAgAGBQJVbicjAAoJENuP0xzK19csvHsH/j1lzkKP4n9ckRWsgm0/TWJw
DhyOSrwq/UXksN4Vty/HlLrwOgfrUtx5/WGOqcAg1b09Cw3rLPbRCQ2aAsulfbj8
wi9i53q5jvuvPAJN+y9nI/96077zLOQgwlfUntWEejnvBk7lqO3erjuvDRObncIp
JS6DQ/DIc0TfxS1KSJJaxMj6Lruqr2xnF7sMLO5SVJzo1LaseTblbVXAasSfwJAi
oRuSHIlObBvmw0lBn49ZJ2c180ToEwShlFLyyg2D0lneyHRUqbz4/IQYK/x89bkD
wSiSPSqN4ujQA/HzUnb84gYWQpnrHsY7POsrlHx/+PzrhXFd8cIq6ISc1TPAaFs=
=1xfC
-----END PGP SIGNATURE-----
je.e...@web.de
I assigned device 04:00.0 and device 04:00.2 to the sys-net VM. But than
I got the error that the sys-net VM can not start because the qrexec
daemon was not able to start.
04:00.0 is an sd card reader with bad driver support in Linux.
I also installed Qubes R3 RC1 again and after the installation I got a
long error message that sys-net VM was not able to start, because of the
qrexec and libvirt. Furthermore I got a message that the disposable VM
could not be created either.
Remark: I don't have Intel VT-d on my laptop.
Best regards
J. Eppler
Alex
Hash: SHA256
On 06/13/2015 12:57 PM, je.e...@web.de wrote:
> Hello,
>
> I assigned device 04:00.0 and device 04:00.2 to the sys-net VM. But
> than I got the error that the sys-net VM can not start because the
> qrexec daemon was not able to start.
>
> 04:00.0 is an sd card reader with bad driver support in Linux.
ethernet on 04:00.1. The SD card had to be passed to the netvm,
otherwise "PCI Reset error" would appear in Qubes Manager, but if
passed the netvm would not successfully boot.
I made a small addition to bashrc, that sudo-launches this script:
- -----------------------
#!/bin/bash
if [ -f /sys/bus/pci/devices/0000\:04\:00.0/remove ]
then
echo -n "1" > /sys/bus/pci/devices/0000\:04\:00.0/remove
fi
- ---------------------
so when I boot qubes and log in as administrator (default user), this
script is executed by some shell (I still don't understand why
sometimes I have to manually open a terminal on dom0 before starting
netvm - sometimes I don't have to, and the device is already correctly
removed, as if some bash was started in the background).
Once the script is run, the sd card reader is "removed" and the netvm
can be configured to receive only the 04:00.1 ethernet card and can
boot successfully.
YMMV
- --
Alex
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2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=uOqS
-----END PGP SIGNATURE-----
je.e...@web.de
this solution works.
Thank you very much.
Clark Venable
I have a new System76 Lemur laptop which I have installed Qubes 3.1 onto. As described in this thread, when I try to launch any VM I receive an error that says;
"Error starting VM: internal error: Unable to reset PCI device 0000:01:00.1: internal error Active 0000:01:00.0 on bus with 0000:01:00.1, not doing reset"
I know from VMSettings-->Devices that
01:00.0 is "Unassigned class [ff00] Realtek Semiconductor Co, Ltd. Device 5287 (rev01)"
and
01:00.1 is "Ethernet controller: Realtek Semiconductor Co, Ltd. RTL81118168-8411 PCI Express Gigabit Ethernet Controller (rev 12)"
So would I edit bashrc to include the following?
if [ -f /sys/bus/pci/devices/0000\:01\:00.0/remove ]
then
echo -n "1" > /sys/bus/pci/devices/0000\:01\:00.0/remove
fi
Thanks for any pointers. New to Qubes and Linux (OS X user since the beta).
Clark
J. Eppler
you can try the following. Assign the ethernet controller to a net VM (e. g. sys-net).
Open a terminal (Konsole, XFCE) and typte in:
sudo echo -n "1" > /sys/bus/pci/devices/0000\:01\:00.0/remove
try to start the net VM, which has the ethernet controller attached.
This should work.
The next step to remove it on startup via script.
J. Eppler
Clark Venable
When I go there for the VM sys-net that controller is already in the 'Selected' pane on the right.
I then right-clicked the sys-net VM in VM Manager and chose to run a command as above.
Got a pop up error which restated prior error.
Clark Venable
Same thing in the dom0 Konsole.
To clarify, should I be typing the escape backstroke character in the command line?
Clark Venable
J. Eppler
> Progress!
great.
> Now a different error message appears when I try to launch a copy of the sys-net template: "Cannot execute qrexec-daemon!"
okay, did you copy the sys-net template or the VM? Can you explain in more detail what you did?
Clark Venable
I'm seeing that I made a mistake by wading into this Qubes pond. I've spent the last few hours with a brand new laptop that boots into qubes and launches dom0, net-vm and firewall-vm, but can't get any wired internet access in the Fedora23 template. And that's not even talking about wireless.
I'm just going to go back to Ubuntu as my first step away from OS X and try Qubes again later.
Thank you *very* much for your help.
Clark
J. Eppler
it would be helpful to know the brand and model of your laptop. Maybe other people run into a similar issue or solve the issue.
best regards
J. Eppler
Clark Venable
BIOS 5.11
Xen: 4.6.0
Kernel 4.1.13-9
Intel Core i3-6100U @ 2.30GHz
Chipset: Sky Lake.....
VGA: Sky Lake integrated Graphics....
Net:
Realtek Semiconductor Co.... RTL8111/8168/8411 PCI Express Gigabit Ethernet
Intel Corporation Wireless 3165 (rev 81)
SCSI: Samsung SSD 850 Rev: 1B6Q
HVM: Active
I/O MMU: Active
TPM: Device not found
Clark Venable
After a night's sleep and some more reading I found a thread which suggested removing the offending PCI device from the "Selected" panel in sys-net-->VM Settings-->Devices, leaving only my Intel Wireless controller available.
After a restart, when I selected the Networking icon from the system tray I was able to see wireless networks and successfully joined one. I next launchd disposable VM and was able to load the fedoraproject.org page and now have network access in all vm's.
